Cyclodextrin-Based Nanostructure Efficiently Delivers siRNA to Glioblastoma Cells Preferentially via Macropinocytosis

International Journal of Molecular Sciences Article Cyclodextrin-Based Nanostructure Efficiently Delivers siRNA to Glioblastoma Cells Preferentially via Macropinocytosis Darío Manzanares 1,2 , María Dolores Pérez-Carrión1,2, JoséLuis Jiménez Blanco 3, Carmen Ortiz Mellet 3, JoséManuel García Fernández 4and Valentín Ceña 1,2,* 1Unidad Asociada Neurodeath, Universidad de Castilla-La Mancha, 02006 Albacete, Spain; [email protected] (D.M.); mdolores.per[email protected] (M.D.P.-C.) 2CIBERNED, Instituto de Salud Carlos III, 28029 Madrid, Spain 3Departamento de Química Orgánica, Facultad de Química, Universidad de Sevilla, 41012 Sevilla, Spain; [email protected] (J.L.J.B.); [email protected] (C.O.M.) 4 Instituto de Investigaciones Qu í micas (IIQ), Consejo Superior de Investigaciones Cient í ficas-Universidad de Sevilla, 41092 Sevilla, Spain, [email protected] *Correspondence: [email protected] Received: 2 November 2020; Accepted: 3 December 2020; Published: 6 December 2020   Abstract: Small interfering ribonucleic acid (siRNA) has the potential to revolutionize therapeutics since it can knockdown very efficiently the target protein. It is starting to be widely used to interfere with cell infection by HIV. However, naked siRNAs are unable to get into the cell, requiring the use of carriers to protect them from degradation and transporting them across the cell membrane. There is no information about which is the most efficient endocytosis route for high siRNA transfection efficiency. One of the most promising carriers to efficiently deliver siRNA are cyclodextrin derivatives. We have used nanocomplexes composed of siRNA and a β -cyclodextrin derivative, AMC6, with a very high transfection efficiency to selectively knockdown clathrin heavy chain, caveolin 1, and p21 Activated Kinase 1 to specifically block clathrin-mediated, caveolin-mediated and macropinocytosis endocytic pathways. The main objective was to identify whether there is a preferential endocytic pathway associated with high siRNA transfection efficiency. We have found that macropinocytosis is the preferential entry pathway for the nanoparticle and its associated siRNA cargo. However, blockade of macropinocytosis does not affect AMC6-mediated transfection efficiency, suggesting that macropinocytosis blockade can be functionally compensated by an increase in clathrin- and caveolin-mediated endocytosis. Keywords: siRNA; β-cyclodextrin; endocytosis; transfection efficiency; micropinocytosis 1. Introduction RNAi represents an effective mechanism that operates in most eukaryotic cells [ 1 ] regulating the activity of miRNAs and so being involved in the regulation of cellular metabolism, replication or malignant transformation [ 2 ]. RNAi has a high potential to be useful in therapeutics since it is able to knockdown proteins involved in the pathogenesis of different diseases [ 3 ], including HIV [ 4 ], by targeting cellular mRNA [5]. Clinical trials have already shown RNAi effectivity in some diseases such as hereditary transthyretin amyloidosis [ 6 ] and acute intermittent porphyria [ 7 ]. In addition, RNAi-mediated knockdown of proteins involved in cancer cell survival has been proposed to potentiate antitumoral actions of drugs, suggesting a new potential therapeutic approach for cancer treatment [ 8 ]. SeveralexogenousactivatorsoftheRNAisystemcanknockdownspecificsequencesinvolvedincellular signaling, small interfering RNAs (siRNAs) being the most widely used [ 9 ]. siRNAs are double-strand Int. J. Mol. Sci. 2020,21, 9306; doi:10.3390/ijms21239306 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2020,21, 9306 2 of 17 RNAs each strand formed by about 21 nucleotides that degrade homologous mRNAs [ 10 ]. siRNAs can knockdown very efficiently the proteins encoded by the target mRNAs, but naked siRNAs are unable to get into the cell, requiring the use of carriers capable of forming nanostructures (nanoparticles; NPs), protecting the siRNA molecules from degradation and transporting them across the cell membrane [11]. siRNA-loaded NPs can gain access to the cell interior through simple diffusion or translocation, an energy-independent process that relies on the NP concentration gradient, but also on other factors such as their size, shape, surface charge or lipophilicity [ 12 ]. However, the most common mechanism used by NPs to enter the cells is endocytosis [ 13 ]. Several endocytosis mechanisms have been described, but the major pathways for endocytic uptake followed by NPs and their cargo to enter the cell are macropinocytosis, clathrin-mediated endocytosis (CME), and caveolin-mediated endocytosis (CVME) [14]. Detailed knowledge of the specific uptake route followed by a given NP is very relevant, since it determines the NP fate as well as the fate of the corresponding cargo, conditioning its possible use as drug delivery system/pharmaceutics. This is even more relevant in the case of siRNA therapeutics since one key step for their efficiency is endosomal escape [ 15 ]. In this sense, CVME is generally advantageous since it seems to avoid the endo-lysosomal system in some cell types [ 16 ]. Nevertheless, other authors report that macropinosomes are more likely to release their content without lysosomal degradation and thus are preferable for siRNA delivery [17]. For a long time, the study of the different endocytosis pathways has been carried out using different chemical inhibitors to block several endocytic pathways [ 18 – 20 ]. However, despite their wide utilization, these compounds are not specific and simultaneously affect several routes [ 21 ], which does not allow a precise dissection of the contribution of a given endocytic pathway to siRNA transfection efficiency. A more specific approach would involve knocking down, using specific siRNAs, certain proteins that play essential roles in the different endocytic uptake routes used by NPs. Thus, we have selectively knocked down clathrin heavy chain (CLTC), an integral component of the clathrin triskelion that forms the clathrin coat in CME [ 22 ] to block the clathrin-dependent pathway [ 23 ]; caveolin 1 (CAV1) to block the caveolin-dependent pathway [ 23 ]; and p21 (Rac1) Activated Kinase 1 (PAK1) to block the macropinocytosis pathway, because its knockdown interrupts the route of activation of macropinocytosis after activation of both Rac1 and Cdc42, which through PAK1 induce the polymerization of actin [24]. As mentioned above, NPs are required to both protect siRNA from degradation and to deliver it to the cells [ 11 ]. To achieve this, we have chosen as carrier a β -cyclodextrin ( β CD)-based molecular multivalent amphiphile [ 25 ] termed AMC6 (Figure 1). AMC6 is a member of the polycationic amphiphilic cyclodextrin nonviral vector family [ 26 ], characterized by a Janus-type architecture with segregated polycationic and lipophilic domains [ 27 ]. It consists of a basket-shaped β CD core with 14 hexanoyl chains at the secondary face and seven tetraethyleneimine branches at the opposite primary rim. These branches at the seven primary positions contain a triazol linker and four protonable nitrogens each, which give to the molecule a total of 28 cationizable centers [ 28 ]. AMC6 forms self-assembled nanoparticles in the presence of siRNA with an estimated size determined by Dynamic Light Scattering of about 110 nm. AMC6 nanoparticles have previously shown no toxicity and a high transfection efficiency in different tumoral cell lines including human U87 and rat C6 glioblastoma [29]. The results presented here show that AMC6 is an excellent siRNA transfection vector that is able to deliver siRNA to decrease target protein levels to about 15 to 25% of control values, a transfection efficiency than is better than that observed in the case of cationic lipid nanoparticles or calcium phosphate nanoparticles, allowing lack-of-function studies. AMC6-mediated siRNA delivery can knockdown key proteins (CLTC, CAV1, and PAK1) of the major endocytic pathways (CME, CVME and macropinocytosis) leading to a functional blockade of them as seen by the marked decrease in the uptake of molecules such as transferrin and 70 K dextran that are specifically taken up by those pathways. This has allowed the dissection of the contribution of each endocytic pathway to AMC6-mediated siRNA delivery to glioblastoma cells. We have found that the major endocytic pathway in T98G glioblastoma cells is macropinocytosis, although the three major pathways contribute to AMC6 efficient siRNA transfection functionally compensating the blockade of one of them. Int. J. Mol. Sci. 2020,21, 9306 3 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 3 of 17 Figure 1. (A) Chemical structure of the polycationic amphiphilic β-cyclodextrin derivative AMC6; (B) Scheme showing siRNA binding to an AMC6 molecule. The results presented here show that AMC6 is an excellent siRNA transfection vector that is able to deliver siRNA to decrease target protein levels to about 15 to 25% of control values, a transfection efficiency than is better than that observed in the case of cationic lipid nanoparticles or calcium phosphate nanoparticles, allowing lack-of-function studies. AMC6-mediated siRNA delivery can knockdown key proteins (CLTC, CAV1, and PAK1) of the major endocytic pathways (CME, CVME and macropinocytosis) leading to a functional blockade of them as seen by the marked decrease in the uptake of molecules such as transferrin and 70 K dextran that are specifically taken up by those pathways. This has allowed the dissection of the contribution of each endocytic pathway to AMC6- mediated siRNA delivery to glioblastoma cells. We have found that the major endocytic pathway in T98G glioblastoma cells is macropinocytosis, although the three major pathways contribute to AMC6 efficient siRNA transfection functionally compensating the blockade of one of them. 2. Results 2.1. Time-Course of siRNA Uptake and Cell Toxicity Incubation of C6, GL261, U87, and T98G glioblastoma cell lines with AMC6/Alexa 488 FluorsiRNA for different times showed a progressive increase in fluorescence up to 6 h (Figure 2). Then, intracellular fluorescence signal was stable up to 8 (C6 cells) or 12 h (GL261 and T98G). In the case of U87 the fluorescent signal remained stable up to 24 h (Figure 2). Raw fluorescence images can be observed in the Supplementary Information (Figure S1). Since all glioma/glioblastoma cell lines showed a similar behavior on Alexa 488 Fluor-siRNA uptake, we decided to perform all the experiments in T98G human glioblastoma cells. This cell line is the only one among the indicated cells that expresses the DNA-repairing enzyme O6-methylguanine–DNA methyltransferase (MGMT) that removes the alkylation caused by temozolomide, a first-line key drug in glioblastoma treatment [30]. The methylation state of the MGMT promoter is a prognostic factor of the patients’ response to this drug [31]. Figure 1. ( A ) Chemical structure of the polycationic amphiphilic β -cyclodextrin derivative AMC6; (B) Scheme showing siRNA binding to an AMC6 molecule. 2. Results 2.1. Time-Course of siRNA Uptake and Cell Toxicity Incubation of C6, GL261, U87, and T98G glioblastoma cell lines with AMC6/Alexa 488 Fluor-siRNA for different times showed a progressive increase in fluorescence up to 6 h (Figure 2). Then, intracellular fluorescence signal was stable up to 8 (C6 cells) or 12 h (GL261 and T98G). In the case of U87 the fluorescent signal remained stable up to 24 h (Figure 2). Raw fluorescence images can be observed in the Supplementary Information (Figure S1). Since all glioma/glioblastoma cell lines showed a similar behavior on Alexa 488 Fluor-siRNA uptake, we decided to perform all the experiments in T98G human glioblastoma cells. This cell line is the only one among the indicated cells that expresses the DNA-repairing enzyme O 6 -methylguanine–DNA methyltransferase (MGMT) that removes the alkylation caused by temozolomide, a first-line key drug in glioblastoma treatment[ 30 ]. The methylation state of the MGMT promoter is a prognostic factor of the patients’ response to this drug [31]. Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 4 of 17 Figure 2. Time-courses for uptake of AMC6/Alexa Fluor 488-labeled siRNA complexes. C6 (A), U87 (B), GL261 (C) and T98G (D) cell lines were treated with AMC6/Alexa Fluor 488-labeled siRNA complexes (1 µM/100 nM) for 2, 4, 6, 8, 10, 12 and 24 h. Fluorescence intensity inside the cells was determined by selecting regions of interest. Data are expressed as the average ± SEM (n = 50). Next, we studied whether the treatment with either the AMC6 vector alone or the nanoplexes was toxic for T98G glioblastoma cells. As it can be observed in Figure 3, AMC6 was not toxic by itself while nanoplexes formed by AMC6 plus either scramble (SCR) siRNA or plus one of the specific siRNAs only slightly increased toxicity in T98G cells (Figure 3). When SCR siRNA at a concentration of 50 nM was coupled to AMC6, LDH released amounted to 11.2 + 1.76% (n = 13). Figure 3. Effect of AMC6/custom siRNA against CLTC, CAV1 and PAK1 on cell viability. Toxicity is expressed as percentage of lactate dehydrogenase (LDH) released to the culture medium. Cells were treated with vehicle (Control), AMC6 (1 µM), AMC6 (1 µM) complexed with non-silencing scramble (SCR) siRNA (100 nM), and AMC6 (1 µM) complexed to 100 nM specific siRNA for CLTC, CAV1, and PAK1. Data are expressed as the mean ± SEM of 8 to 20 experiments. 2.2. Knockdown of Key Proteins Involved in Endocytic Pathways We decided to study the effect of increasing concentrations of specific siRNA aimed to knockdown key proteins involved in CME (CLCT), CVME (CAV1), and macropinocytosis (PAK1). As it can be observed in Figure 4, neither AMC alone nor the AMC6/SCR siRNA (100 nM) nanoplexes significantly modified the target protein levels after 72 h of treatment (Figure 4). Figure 2. Time-courses for uptake of AMC6/Alexa Fluor 488-labeled siRNA complexes. C6 ( A ), U87 ( B ), GL261 ( C ) and T98G ( D ) cell lines were treated with AMC6/Alexa Fluor 488-labeled siRNA complexes (1 µ M/100 nM) for 2, 4, 6, 8, 10, 12 and 24 h. Fluorescence intensity inside the cells was determined by selecting regions of interest. Data are expressed as the average ±SEM (n=50). Int. J. Mol. Sci. 2020,21, 9306 4 of 17 Next, we studied whether the treatment with either the AMC6 vector alone or the nanoplexes was toxic for T98G glioblastoma cells. As it can be observed in Figure 3, AMC6 was not toxic by itself while nanoplexes formed by AMC6 plus either scramble (SCR) siRNA or plus one of the specific siRNAs only slightly increased toxicity in T98G cells (Figure 3). When SCR siRNA at a concentration of 50 nM was coupled to AMC6, LDH released amounted to 11.2 +1.76% (n=13). Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 4 of 17 Figure 2. Time-courses for uptake of AMC6/Alexa Fluor 488-labeled siRNA complexes. C6 (A), U87 (B), GL261 (C) and T98G (D) cell lines were treated with AMC6/Alexa Fluor 488-labeled siRNA complexes (1 µM/100 nM) for 2, 4, 6, 8, 10, 12 and 24 h. Fluorescence intensity inside the cells was determined by selecting regions of interest. Data are expressed as the average ± SEM (n = 50). Next, we studied whether the treatment with either the AMC6 vector alone or the nanoplexes was toxic for T98G glioblastoma cells. As it can be observed in Figure 3, AMC6 was not toxic by itself while nanoplexes formed by AMC6 plus either scramble (SCR) siRNA or plus one of the specific siRNAs only slightly increased toxicity in T98G cells (Figure 3). When SCR siRNA at a concentration of 50 nM was coupled to AMC6, LDH released amounted to 11.2 + 1.76% (n = 13). Figure 3. Effect of AMC6/custom siRNA against CLTC, CAV1 and PAK1 on cell viability. Toxicity is expressed as percentage of lactate dehydrogenase (LDH) released to the culture medium. Cells were treated with vehicle (Control), AMC6 (1 µM), AMC6 (1 µM) complexed with non-silencing scramble (SCR) siRNA (100 nM), and AMC6 (1 µM) complexed to 100 nM specific siRNA for CLTC, CAV1, and PAK1. Data are expressed as the mean ± SEM of 8 to 20 experiments. 2.2. Knockdown of Key Proteins Involved in Endocytic Pathways We decided to study the effect of increasing concentrations of specific siRNA aimed to knockdown key proteins involved in CME (CLCT), CVME (CAV1), and macropinocytosis (PAK1). As it can be observed in Figure 4, neither AMC alone nor the AMC6/SCR siRNA (100 nM) nanoplexes significantly modified the target protein levels after 72 h of treatment (Figure 4). Figure 3. Effect of AMC6/custom siRNA against CLTC, CAV1 and PAK1 on cell viability. Toxicity is expressed as percentage of lactate dehydrogenase (LDH) released to the culture medium. Cells were treated with vehicle (Control), AMC6 (1 µ M), AMC6 (1 µ M) complexed with non-silencing scramble (SCR) siRNA (100 nM), and AMC6 (1 µ M) complexed to 100 nM specific siRNA for CLTC, CAV1, and PAK1. Data are expressed as the mean ±SEM of 8 to 20 experiments. 2.2. Knockdown of Key Proteins Involved in Endocytic Pathways We decided to study the effect of increasing concentrations of specific siRNA aimed to knockdown key proteins involved in CME (CLCT), CVME (CAV1), and macropinocytosis (PAK1). As it can be observed in Figure 4, neither AMC alone nor the AMC6/SCR siRNA (100 nM) nanoplexes significantly modified the target protein levels after 72 h of treatment (Figure 4). Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 5 of 17 Figure 4. Knockdown of CLTC, CAV1 and PAK1 proteins by specific siRNAs. Cells were treated for 72 h with vehicle (Control), AMC6 (1 µM), AMC6 (1 µM) complexed with SCR siRNA (100 nM) or AMC6 (1 µM) complexed to increasing concentrations of specific siRNAs (25, 50 and 100 nM). The graph shows the ratios between the optical densities of CLTC and α-tubulin (A), CAV1 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (B) and PAK1 and GAPDH (C), expressed as percentage in comparison to the vehicle (control). Western blot images belong to a representative experiment. Data are expressed as the mean ± SEM of 3 individual experiments. ** p < 0.01; *** p < 0.001 as compared to control. In contrast, nanoplexes formed using specific siRNA at 50 and 100 nM concentrations knocked down the target protein levels by more than 80% (Figure 4). From these data, 50 nM siRNA was selected as the lowest siRNA concentration providing a very significant (more than 80%) reduction in target protein levels, which would allow studying the effect of lack-of-function for the target protein on the cell uptake efficiency of the nanocomplexes. No specific siRNA-mediated cross inhibition was observed with the other two proteins studied (supporting information, Figure S2). As it can be observed in Figure 5, a significant protein knockdown, allowing lack-of-function studies, could be achieved at 48 h for the three studied proteins. Thus, this time point was also selected for future experiments. Figure 5. Time-course of AMC6/custom siRNA against CLTC, CAV1 and PAK1. Cells were treated with vehicle (control) for 72 h, and AMC6 (1 µM) complexed to CLTC (A), CAV1 (B) and PAK1 (C) targeting siRNA (50 nM) for 24, 48 and 72 h; The graph shows the ratios between the optical densities of CLTC and α-tubulin (A) CAV1 and GAPDH (B) and PAK1 and GAPDH (C), expressed as percentage in comparison to the vehicle (control). Western blot images belong to a representative experiment. Data are expressed as the mean ± SEM of 3 individual experiments. ** p < 0.01; *** p < 0.001 as compared to control. Figure 4. Knockdown of CLTC, CAV1 and PAK1 proteins by specific siRNAs. Cells were treated for 72 h with vehicle (Control), AMC6 (1 µ M), AMC6 (1 µ M) complexed with SCR siRNA (100 nM) or AMC6 (1 µ M) complexed to increasing concentrations of specific siRNAs (25, 50 and 100 nM). The graph shows the ratios between the optical densities of CLTC and α -tubulin ( A ), CAV1 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) ( B ) and PAK1 and GAPDH ( C ), expressed as percentage in comparison to the vehicle (control). Western blot images belong to a representative experiment. Data are expressed as the mean ± SEM of 3 individual experiments. ** p<0.01; *** p<0.001 as compared to control. Int. J. Mol. Sci. 2020,21, 9306 5 of 17 In contrast, nanoplexes formed using specific siRNA at 50 and 100 nM concentrations knocked down the target protein levels by more than 80% (Figure 4). From these data, 50 nM siRNA was selected as the lowest siRNA concentration providing a very significant (more than 80%) reduction in target protein levels, which would allow studying the effect of lack-of-function for the target protein on the cell uptake efficiency of the nanocomplexes. No specific siRNA-mediated cross inhibition was observed with the other two proteins studied (supporting information, Figure S2). As it can be observed in Figure 5, a significant protein knockdown, allowing lack-of-function studies, could be achieved at 48 h for the three studied proteins. Thus, this time point was also selected for future experiments. Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 5 of 17 Figure 4. Knockdown of CLTC, CAV1 and PAK1 proteins by specific siRNAs. Cells were treated for 72 h with vehicle (Control), AMC6 (1 µM), AMC6 (1 µM) complexed with SCR siRNA (100 nM) or AMC6 (1 µM) complexed to increasing concentrations of specific siRNAs (25, 50 and 100 nM). The graph shows the ratios between the optical densities of CLTC and α-tubulin (A), CAV1 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (B) and PAK1 and GAPDH (C), expressed as percentage in comparison to the vehicle (control). Western blot images belong to a representative experiment. Data are expressed as the mean ± SEM of 3 individual experiments. ** p < 0.01; *** p < 0.001 as compared to control. In contrast, nanoplexes formed using specific siRNA at 50 and 100 nM concentrations knocked down the target protein levels by more than 80% (Figure 4). From these data, 50 nM siRNA was selected as the lowest siRNA concentration providing a very significant (more than 80%) reduction in target protein levels, which would allow studying the effect of lack-of-function for the target protein on the cell uptake efficiency of the nanocomplexes. No specific siRNA-mediated cross inhibition was observed with the other two proteins studied (supporting information, Figure S2). As it can be observed in Figure 5, a significant protein knockdown, allowing lack-of-function studies, could be achieved at 48 h for the three studied proteins. Thus, this time point was also selected for future experiments. Figure 5. Time-course of AMC6/custom siRNA against CLTC, CAV1 and PAK1. Cells were treated with vehicle (control) for 72 h, and AMC6 (1 µM) complexed to CLTC (A), CAV1 (B) and PAK1 (C) targeting siRNA (50 nM) for 24, 48 and 72 h; The graph shows the ratios between the optical densities of CLTC and α-tubulin (A) CAV1 and GAPDH (B) and PAK1 and GAPDH (C), expressed as percentage in comparison to the vehicle (control). Western blot images belong to a representative experiment. Data are expressed as the mean ± SEM of 3 individual experiments. ** p < 0.01; *** p < 0.001 as compared to control. Figure 5. Time-course of AMC6/custom siRNA against CLTC, CAV1 and PAK1. Cells were treated with vehicle (control) for 72 h, and AMC6 (1 µ M) complexed to CLTC ( A ), CAV1 ( B ) and PAK1 ( C ) targeting siRNA (50 nM) for 24, 48 and 72 h; The graph shows the ratios between the optical densities of CLTC and α -tubulin ( A ) CAV1 and GAPDH ( B ) and PAK1 and GAPDH ( C ), expressed as percentage in comparison to the vehicle (control). Western blot images belong to a representative experiment. Data are expressed as the mean ±SEM of 3 individual experiments. ** p<0.01; *** p<0.001 as compared to control. 2.3. Effect of Signaling Protein Knockdown on Function of Endocytic Mechanisms Once optimal conditions for efficient knockdown of the target proteins were established, we tested whether removal of the function of these proteins was able to block endocytosis through the involved endocytic pathway. To explore this, we used Alexa 555-labelled human transferrin that is taken up selectively by a CME [ 32 ] and tetramethylrhodamine (TMR)-labelled 70 kDa dextran which is specifically endocytosed by macropinocytosis [ 33 ]. No compound has been described yet to be specifically taken by the caveolin-dependent pathway, although cholera toxin is taken preferentially, but not exclusively by this pathway [34]. As it can be observed in Figure 6, knockdown of CLTC for 48 h decreased transferrin uptake by about 70%, while nanoplexes formed by AMC6 and SCR siRNA (50 nM) lacked any effect. Moreover, when PAK1 was knocked down, TMR-labeled 70 kDa dextran uptake was decreased by almost 60% (Figure 7). In this case, nanoplexes formed by AMC6 and SCR slightly decreased TMR-70 kDa dextran uptake (about 15%) indicating a small unspecific effect of siRNA on this endocytic pathway. 2.4. Effect of Selective Blockade of Endocytic Pathways on AMC6/siRNA Nanoplexes Uptake To study whether the selective blockade of a single endocytic pathway (CME, CVME or macropinocytosis) could lead to a marked decrease in AMC6-mediated FAM-siRNA uptake, we treated T98G glioblastoma cells with nanoplexes formed by AMC6 and siRNA, either SCR or aimed to knockdown CLTC, CAV1 or PAK1, for 48 h. At that time, we washed the cells with culture medium to remove non-taken up nanoplexes. Then, new nanoplexes formed by AMC6 and Alexa 488 Fluor-siRNA were added to assess siRNA uptake. Int. J. Mol. Sci. 2020,21, 9306 6 of 17 As it can be observed in Figure 8, only knocking down PAK1, implying macropinocytosis inhibition (Figure 7), reduced the cellular uptake of Alexa 488 Fluor-siRNA by 35%, suggesting that the non-blocked uptake pathways cannot fully compensate for macropinocytosis blockade. On the other hand, knocking down either CLTC or CAV1 did not decrease AMC6-mediated Alexa 488 Fluor-siRNA uptake, suggesting that either these endocytic routes are not significantly involved in the internalization of the nanocomplexes or that they can be compensated by the other non-blocked endocytic pathways. Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 6 of 17 2.3. Effect of Signaling Protein Knockdown on Function of Endocytic Mechanisms Once optimal conditions for efficient knockdown of the target proteins were established, we tested whether removal of the function of these proteins was able to block endocytosis through the involved endocytic pathway. To explore this, we used Alexa 555-labelled human transferrin that is taken up selectively by a CME [32] and tetramethylrhodamine (TMR)-labelled 70 kDa dextran which is specifically endocytosed by macropinocytosis [33]. No compound has been described yet to be specifically taken by the caveolin-dependent pathway, although cholera toxin is taken preferentially, but not exclusively by this pathway [34]. As it can be observed in Figure 6, knockdown of CLTC for 48 h decreased transferrin uptake by about 70%, while nanoplexes formed by AMC6 and SCR siRNA (50 nM) lacked any effect. Moreover, when PAK1 was knocked down, TMR-labeled 70 kDa dextran uptake was decreased by almost 60% (Figure 7). In this case, nanoplexes formed by AMC6 and SCR slightly decreased TMR-70 kDa dextran uptake (about 15%) indicating a small unspecific effect of siRNA on this endocytic pathway. Figure 6. Effect of the downregulation of CLTC on the uptake of Alexa Fluor 555 conjugated transferrin. T98G cells were pre-treated with vehicle (control), AMC6/SCR siRNA and AMC6/anti- CLTC siRNA complexes (both 1 µM/50 nM) for 48 h to be later treated for 30 min with fluorescent Alexa Fluor 555 labeled human transferrin (25 µg/mL). The images panel (A) shows Alexa Fluor 555- transferrin (red), Hoescht 33,342 (blue) and the overlay of both images. In addition, differential interference contrast (DIC) images are also shown. Scale bar: 25 µm; (B) Quantification of fluorescence (expressed in Relative Fluorescence Units; RFU) for the 3 experimental conditions. Data are expressed as the mean ± SEM (n = 50). *** p < 0.001, compared to control cells. Figure 6. Effect of the downregulation of CLTC on the uptake of Alexa Fluor 555 conjugated transferrin. T98G cells were pre-treated with vehicle (control), AMC6/SCR siRNA and AMC6/anti-CLTC siRNA complexes (both 1 µ M/50 nM) for 48 h to be later treated for 30 min with fluorescent Alexa Fluor 555 labeled human transferrin (25 µ g/mL). The images panel ( A ) shows Alexa Fluor 555-transferrin (red), Hoescht 33,342 (blue) and the overlay of both images. In addition, differential interference contrast (DIC) images are also shown. Scale bar: 25 µ m; ( B ) Quantification of fluorescence (expressed in Relative Fluorescence Units; RFU) for the 3 experimental conditions. Data are expressed as the mean ± SEM (n=50). *** p<0.001, compared to control cells. 2.5. Effect of Specific Blockade of siRNA Uptake Pathways on Transfection Efficiency Once we established that we could selectively and efficiently block either CME, CVME or macropinocytosis, we explored whether the decrease in AMC6-mediated Alexa Fluor 488-labelled siRNA cellular uptake following the observed knockdown of PAK1 and macropinocytosis blockade was correlated with a decrease in AMC6 transfection ability. For this purpose, we first selectively Int. J. Mol. Sci. 2020,21, 9306 7 of 17 knocked down either PAK1, CLTC or CAV1 (the latter used as controls) by incubating the cells with 50 nM of the specific siRNA (50 nM) for 48 h. At that time, target proteins levels were decreased to about 20% of their initial values (Figure 5) and the corresponding pathways blocked (Figures 6and 7). Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 7 of 17 Figure 7. Effect of the downregulation of PAK1 in the uptake of TMR-conjugated 70 kDa dextran. T98G cells were pre-treated with vehicle (control), AMC6/SCR siRNA and AMC6/anti PAK1 siRNA (both 1 µM/50 nM) for 48 h to be later treated for 1 h with TMR-labeled 70 kDa dextran (0.5/mL). The images panel (A) shows TMR-70 kDa dextran (red), Hoescht 33,342 (blue) and the overlay of both images). In addition, differential interference contrast (DIC) images are also shown. Scale bars: 25 µm; (B) Quantification of fluorescence (expressed in RFU) for the 3 experimental conditions. Data are expressed as the mean ± SEM (n = 50).* p < 0.05, *** p < 0.001, compared to control cells. 2.4. Effect of Selective Blockade of Endocytic Pathways on AMC6/siRNA Nanoplexes Uptake To study whether the selective blockade of a single endocytic pathway (CME, CVME or macropinocytosis) could lead to a marked decrease in AMC6-mediated FAM-siRNA uptake, we treated T98G glioblastoma cells with nanoplexes formed by AMC6 and siRNA, either SCR or aimed to knockdown CLTC, CAV1 or PAK1, for 48 h. At that time, we washed the cells with culture medium to remove non-taken up nanoplexes. Then, new nanoplexes formed by AMC6 and Alexa 488 FluorsiRNA were added to assess siRNA uptake. As it can be observed in Figure 8, only knocking down PAK1, implying macropinocytosis inhibition (Figure 7), reduced the cellular uptake of Alexa 488 Fluor-siRNA by 35%, suggesting that the non-blocked uptake pathways cannot fully compensate for macropinocytosis blockade. On the other hand, knocking down either CLTC or CAV1 did not decrease AMC6-mediated Alexa 488 FluorsiRNA uptake, suggesting that either these endocytic routes are not significantly involved in the internalization of the nanocomplexes or that they can be compensated by the other non-blocked endocytic pathways. Figure 7. Effect of the downregulation of PAK1 in the uptake of TMR-conjugated 70 kDa dextran. T98G cells were pre-treated with vehicle (control), AMC6/SCR siRNA and AMC6/anti PAK1 siRNA (both 1 µ M/50 nM) for 48 h to be later treated for 1 h with TMR-labeled 70 kDa dextran (0.5/mL). The images panel ( A ) shows TMR-70 kDa dextran (red), Hoescht 33,342 (blue) and the overlay of both images). In addition, differential interference contrast (DIC) images are also shown. Scale bars: 25 µ m; ( B ) Quantification of fluorescence (expressed in RFU) for the 3 experimental conditions. Data are expressed as the mean ±SEM (n=50).* p<0.05, *** p<0.001, compared to control cells. Then, the medium was changed to remove non-taken nanoplexes and new medium containing nanoplexes formed by either AMC6 (1 µ M) plus either SCR siRNA (50 nM) or anti-p42 mitogen-activated protein kinase (p42 MAPK) siRNA (50 nM) was added and incubation continued for additional 48 h. This anti-p42 MAPK siRNA has shown previously that is able to reduce p42 MAPK protein levels to about 20% of control values [ 29 ]. As it can be observed in Figure 9, blockade of one single pathway does not prevent β CD-based nanoplexes to efficiently knockdown p42 MAPK protein levels, suggesting that, in functional terms, AMC6/siRNA nanocomplexes are taken up by the three major endocytic Int. J. Mol. Sci. 2020,21, 9306 8 of 17 uptake pathways that compensate among them, maintaining a high level of siRNA transfection if one of those mechanisms is blocked. Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 8 of 17 Figure 8. Effect of knocking down key proteins in CME, CVME and macropinocytosis on the uptake of AMC6/Alexa Fluor 488-labelled siRNA. T98G glioblastoma cells were pre-treated for 48 h with vehicle (control) and AMC6/siRNA complexes (1 µM/50 nM) consisting in SCR siRNA, or specific siRNAs targeting CLTC, CAV1, and PAK1. After this, cells were treated for 6 h with AMC6/alexa488 siRNA (green) complexes (1 µM/100 nM) and with Hoechst 33,342 (25µ/mL) (blue) for 10 min before image recording. (A) The images panel (A) shows Alexa Fluor 488-siRNA, Hoescht 33,342 and the overlay of both images). In addition, DIC images are also shown. Scale bars: 25 µm; (B) Quantification of fluorescence (expressed in RFU) for all the experimental conditions. Data are expressed as the mean ± SEM (n = 100). *** p < 0.001, compared to control cells. 2.5. Effect of Specific Blockade of siRNA Uptake Pathways on Transfection Efficiency Once we established that we could selectively and efficiently block either CME, CVME or macropinocytosis, we explored whether the decrease in AMC6-mediated Alexa Fluor 488-labelled siRNA cellular uptake following the observed knockdown of PAK1 and macropinocytosis blockade was correlated with a decrease in AMC6 transfection ability. For this purpose, we first selectively knocked down either PAK1, CLTC or CAV1 (the latter used as controls) by incubating the cells with 50 nM of the specific siRNA (50 nM) for 48 h. At that time, target proteins levels were decreased to Figure 8. Effect of knocking down key proteins in CME, CVME and macropinocytosis on the uptake of AMC6/Alexa Fluor 488-labelled siRNA. T98G glioblastoma cells were pre-treated for 48 h with vehicle (control) and AMC6/siRNA complexes (1 µ M/50 nM) consisting in SCR siRNA, or specific siRNAs targeting CLTC, CAV1, and PAK1. After this, cells were treated for 6 h with AMC6/alexa488 siRNA (green) complexes (1 µ M/100 nM) and with Hoechst 33,342 (25 µ /mL) (blue) for 10 min before image recording. ( A ) The images panel ( A ) shows Alexa Fluor 488-siRNA, Hoescht 33,342 and the overlay of both images). In addition, DIC images are also shown. Scale bars: 25 µ m; ( B ) Quantification of fluorescence (expressed in RFU) for all the experimental conditions. Data are expressed as the mean ± SEM (n=100). *** p<0.001, compared to control cells. Int. J. Mol. Sci. 2020,21, 9306 9 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 9 of 17 about 20% of their initial values (Figure 5) and the corresponding pathways blocked (Figures 6 and 7). Then, the medium was changed to remove non-taken nanoplexes and new medium containing nanoplexes formed by either AMC6 (1 µM) plus either SCR siRNA (50 nM) or anti-p42 mitogenactivated protein kinase (p42 MAPK) siRNA (50 nM) was added and incubation continued for additional 48 h. This anti-p42 MAPK siRNA has shown previously that is able to reduce p42 MAPK protein levels to about 20% of control values [29]. As it can be observed in Figure 9, blockade of one single pathway does not prevent βCD-based nanoplexes to efficiently knockdown p42 MAPK protein levels, suggesting that, in functional terms, AMC6/siRNA nanocomplexes are taken up by the three major endocytic uptake pathways that compensate among them, maintaining a high level of siRNA transfection if one of those mechanisms is blocked. Figure 9. Effect of selective knockdown of CLTC (A,B), CAV1 (C,D), and PAK1(E,F) on AMC6- mediated transfection efficiency. T98G cells were treated for 48 h with either vehicle (control), AMC6 (1 µM) or nanoplexes formed by AMC6 (1 uM)/specific siRNA (50 nM) targeting CLTC (A), CAV1 (C) and PAK1 (E). Then, the culture medium was replaced by one containing nanoplexes formed by AMC6 and 50 nM of either SCR siRNA or anti-p42- mitogen-activated protein kinase (p42 MAPK) Figure 9. Effect of selective knockdown of CLTC ( A , B ), CAV1 ( C , D ), and PAK1( E , F ) on AMC6-mediated transfection efficiency. T98G cells were treated for 48 h with either vehicle (control), AMC6 (1 µ M) or nanoplexes formed by AMC6 (1 uM)/specific siRNA (50 nM) targeting CLTC ( A ), CAV1 ( C ) and PAK1 ( E ). Then, the culture medium was replaced by one containing nanoplexes formed by AMC6 and 50 nM of either SCR siRNA or anti-p42- mitogen-activated protein kinase (p42 MAPK) and cells were incubated for additional 48 h. CLTC ( A ), CAV1 ( C ), and PAK1 ( E ) protein levels were determined respect to control values. In the experiments shown in panels B, D, and F, the cells received the same treatments, but p42-MAPK levels were determined instead. Western blot images shown representative experiments for each protocol. Data are expressed as mean ± SEM of 3 or 4 individual experiments for each group. *** p<0.001 as compared to control; ns, statistically non-significant. Int. J. Mol. Sci. 2020,21, 9306 16 of 17 38. Xu, C.; Wu, Y.L.; Li, Z.; Loh, X.J. Cyclodextrin-based sustained gene release systems: A supramolecular solution towards clinical applications. Mater. Chem. Front. 2019,3, 181–192. [CrossRef] 39. Davis, M.E.; Zuckerman, J.E.; Choi, C.H.; Seligson, D.; Tolcher, A.; Alabi, C.A.; Yen, Y.; Heidel, J.D.; Ribas, A. Evidence of RNAi in humans from systemically administered siRNA via targeted nanoparticles. Nature 2010 , 464, 1067–1070. [CrossRef] 40. Jim é nez Blanco, J.J.; Ortega-Caballero, F.; Blanco-Fern á ndez, L.; Carmona, T.; Marcelo, G.; Mart í nez-Negro, M.; Aicart, E.; Junquera, E.; Mendicuti, F.; Tros de Ilarduya, C.; et al. Trehalose-based Janus cyclooligosaccharides: The “click” synthesis and DNA-directed assembly into pH-sensitive transfectious nanoparticles. Chem. Commun. 2016,52, 10117–10120. [CrossRef] 41. Carbajo-Gordillo, A.I.; Rodr í guez-Lavado, J.; Jim é nez Blanco, J.L.J.; Benito, J.M.; Di Giorgio, C.; V é laz, I.; Tros de Ilarduya, C.; Ortiz Mellet, C.; Garc í a Fern á ndez, J.M. Trehalose-based Siamese twin amphiphiles with tunable self-assembling, DNA nanocomplexing and gene delivery properties. Chem. Commun. 2019 ,55, 8227–8230. [CrossRef] [PubMed] 42. Mercer, J.; Helenius, A. Gulping rather than sipping: Macropinocytosis as a way of virus entry. Curr. Opin. Microbiol. 2012,15, 490–499. [CrossRef] [PubMed] 43. McMahon, H.T.; Boucrot, E. Molecular mechanism and physiological functions of clathrin-mediated endocytosis. Nat. Rev. Mol. Cell Biol. 2011,12, 517–533. [CrossRef] [PubMed] 44. Pelkmans, L.; Burli, T.; Zerial, M.; Helenius, A. Caveolin-stabilized membrane domains as multifunctional transport and sorting devices in endocytic membrane traffic. Cell 2004,118, 767–780. [CrossRef] 45. Foroozandeh, P.; Aziz, A.A. Insight into Cellular Uptake and Intracellular Trafficking of Nanoparticles. Nanoscale Res. Lett. 2018,13, 339. [CrossRef] 46. Chen, D.; Yang, C.; Liu, S.; Hang, W.; Wang, X.; Chen, J.; Shi, A. SAC-1 ensures epithelial endocytic recycling by restricting ARF-6 activity. J. Cell Biol. 2018,217, 2121–2139. [CrossRef] 47. Gauthier-Rouviere, C.; Bodin, S.; Comunale, F.; Planchon, D. Flotillin membrane domains in cancer. Cancer Metastasis Rev. 2020,39, 361–374. [CrossRef] 48. Parton, R.G.; Joggerst, B.; Simons, K. Regulated internalization of caveolae. J. Cell Biol. 1994 ,127, 1199–1215. [CrossRef] 49. West, M.A.; Bretscher, M.S.; Watts, C. Distinct endocytotic pathways in epidermal growth factor-stimulated human carcinoma A431 cells. J. Cell Biol. 1989,109, 2731–2739. [CrossRef] 50. Chang, C.C.; Wu, M.; Yuan, F. Role of specific endocytic pathways in electrotransfection of cells. Mol. Ther. Methods Clin. Dev. 2014,1, 14058. [CrossRef] 51. Billiet, L.; Gomez, J.P.; Berchel, M.; Jaffres, P.A.; Le Gall, T.; Montier, T.; Bertrand, E.; Cheradame, H.; Guegan, P.; Mevel, M.; et al. Gene transfer by chemical vectors, and endocytosis routes of polyplexes, lipoplexes and lipopolyplexes in a myoblast cell line. Biomaterials 2012 ,33, 2980–2990. [CrossRef] [PubMed] 52. Pavlides, S.; Gutierrez-Pajares, J.L.; Iturrieta, J.; Lisanti, M.P.; Frank, P.G. Endothelial caveolin-1 plays a major role in the development of atherosclerosis. Cell Tissue Res. 2014,356, 147–157. [CrossRef] [PubMed] 53. Lee, S.Y. Temozolomide resistance in glioblastoma multiforme. Genes Dis. 2016 ,3, 198–210. [CrossRef] [PubMed] 54. Huang, J.L.; Jiang, G.; Song, Q.X.; Gu, X.; Hu, M.; Wang, X.L.; Song, H.H.; Chen, L.P.; Lin, Y.Y.; Jiang, D.; et al. Lipoprotein-biomimetic nanostructure enables efficient targeting delivery of siRNA to Ras-activated glioblastoma cells via macropinocytosis. Nat. Commun. 2017,8, 15144. [CrossRef] [PubMed] 55. Recouvreux, M.V.; Commisso, C. Macropinocytosis: A Metabolic Adaptation to Nutrient Stress in Cancer. Front. Endocrinol. 2017,8, 261. [CrossRef] [PubMed] 56. Swanson, J.A.; Yoshida, S. Macropinosomes as units of signal transduction. Philos. Trans. R. Soc. B 2019 ,374, 20180157. [CrossRef] 57. Cao, Y.; Wang, X.; Li, Y.; Evers, M.; Zhang, H.; Chen, X. Extracellular and macropinocytosis internalized ATP work together to induce epithelial-mesenchymal transition and other early metastatic activities in lung cancer. Cancer Cell Int. 2019,19, 254. [CrossRef] 58. Iglesias, R.; Koria, P. Leveraging growth factor induced macropinocytosis for targeted treatment of lung cancer. Med. Oncol. 2015,32, 259. [CrossRef] 59. Kankaanpaa, P.; Tiitta, S.; Bergman, L.; Puranen, A.B.; von Haartman, E.; Linden, M.; Heino, J. Cellular recognition and macropinocytosis-like internalization of nanoparticles targeted to integrin alpha2beta1. Nanoscale 2015,7, 17889–17901. [CrossRef] Int. J. Mol. Sci. 2020,21, 9306 17 of 17 60. Schindelin, J.; Arganda-Carreras, I.; Frise, E.; Kaynig, V.; Longair, M.; Pietzsch, T.; Preibisch, S.; Rueden, C.; Saalfeld, S.; Schmid, B.; et al. Fiji: An open-source platform for biological-image analysis. Nat. Methods 2012 , 9, 676–682. [CrossRef] 61. Posadas, I.; P é rez-Mart í nez, F.C.; Guerra, J.; Sanchez-Verd ú , P.; Ceña, V. Cofilin activation mediates Bax translocation to mitochondria during excitotoxic neuronal death. J. Neurochem. 2012 ,120, 515–527. [CrossRef] [PubMed] 62. P é rez-Mart í nez, F.C.; Carri ó n, B.; Lucio, M.I.; Rubio, N.; Herrero, M.A.; V á zquez, E.; Ceña, V. Enhanced docetaxel-mediated cytotoxicity in human prostate cancer cells through knockdown of cofilin-1 by carbon nanohorn delivered siRNA. Biomaterials 2012,33, 8152–8159. [CrossRef] [PubMed] Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).